Detection the topics of Facebook posts using text mining with Latent Dirichlet Allocation (LDA) algorithm
The development of artificial intelligence technologies has led to their massive integration in various fields, including daily life. Text data plays a pivotal role in the world of artificial intelligence, especially in machine learning, allowing valuable insights to be extracted from massive data...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
University of Baghdad
2025-01-01
|
| Series: | Ibn Al-Haitham Journal for Pure and Applied Sciences |
| Subjects: | |
| Online Access: | https://jih.uobaghdad.edu.iq/index.php/j/article/view/4033 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832591994165133312 |
|---|---|
| author | Shahlaa Mashhadani |
| author_facet | Shahlaa Mashhadani |
| author_sort | Shahlaa Mashhadani |
| collection | DOAJ |
| description |
The development of artificial intelligence technologies has led to their massive integration in various fields, including daily life. Text data plays a pivotal role in the world of artificial intelligence, especially in machine learning, allowing valuable insights to be extracted from massive data sets to help make informed decisions. Latent Dirichlet Allocation (LDA) and digital forensics intersect through analyzing and classifying textual digital evidence in social media, including Facebook, in which text data is the main focus. This technique is particularly a useful topic modeling technique for uncovering hidden patterns in text data, which can be particularly useful in digital forensics taken from Facebook, including text analysis and evidence discovery, where LDA is used to extract large amounts of unstructured text data from meaningful topics, such as emails, documents, or chat logs. Investigators often deal with huge amounts of text-based evidence, so this technique helps them identify topics, such as fraud, especially in relation to text data, which is the core of our research. It not only improves effort and time but also carries a huge potential for security packages. This work presents a method for processing Facebook posts with the help of a Latent Dirichlet Allocation (LDA) ruleset to classify these texts into coherent themes. The significance of the research lies in its ability to discover themes within each post, which is crucial for analyzing user behavior and addressing security concerns. The use of relevant Facebook data enhances the real-world relevance of the results, facilitating targeted analysis based on the language patterns used by users in these posts and thus contributing to the success of security objectives. In evaluating existing methodologies, this study demonstrates improved performance by optimizing the LDA ruleset to more accurately match the unique features of the target statistics. This improvement leads to improved performance and reduced errors. The results of this study demonstrate the effectiveness of using the LDA approach, as it showed significant improvements over traditional strategies in terms of accuracy and applicability to real-world security situations and digital analytics.
|
| format | Article |
| id | doaj-art-5705d249717c4baf83f72493cea38f14 |
| institution | Kabale University |
| issn | 1609-4042 2521-3407 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | University of Baghdad |
| record_format | Article |
| series | Ibn Al-Haitham Journal for Pure and Applied Sciences |
| spelling | doaj-art-5705d249717c4baf83f72493cea38f142025-01-22T01:21:02ZengUniversity of BaghdadIbn Al-Haitham Journal for Pure and Applied Sciences1609-40422521-34072025-01-0138110.30526/38.1.4033Detection the topics of Facebook posts using text mining with Latent Dirichlet Allocation (LDA) algorithmShahlaa Mashhadani0https://orcid.org/0000-0003-2193-4796Computer Science Department, College of Education for Pure Sciences (Ibn Al-Haitham), University of Baghdad, Baghdad, Iraq. The development of artificial intelligence technologies has led to their massive integration in various fields, including daily life. Text data plays a pivotal role in the world of artificial intelligence, especially in machine learning, allowing valuable insights to be extracted from massive data sets to help make informed decisions. Latent Dirichlet Allocation (LDA) and digital forensics intersect through analyzing and classifying textual digital evidence in social media, including Facebook, in which text data is the main focus. This technique is particularly a useful topic modeling technique for uncovering hidden patterns in text data, which can be particularly useful in digital forensics taken from Facebook, including text analysis and evidence discovery, where LDA is used to extract large amounts of unstructured text data from meaningful topics, such as emails, documents, or chat logs. Investigators often deal with huge amounts of text-based evidence, so this technique helps them identify topics, such as fraud, especially in relation to text data, which is the core of our research. It not only improves effort and time but also carries a huge potential for security packages. This work presents a method for processing Facebook posts with the help of a Latent Dirichlet Allocation (LDA) ruleset to classify these texts into coherent themes. The significance of the research lies in its ability to discover themes within each post, which is crucial for analyzing user behavior and addressing security concerns. The use of relevant Facebook data enhances the real-world relevance of the results, facilitating targeted analysis based on the language patterns used by users in these posts and thus contributing to the success of security objectives. In evaluating existing methodologies, this study demonstrates improved performance by optimizing the LDA ruleset to more accurately match the unique features of the target statistics. This improvement leads to improved performance and reduced errors. The results of this study demonstrate the effectiveness of using the LDA approach, as it showed significant improvements over traditional strategies in terms of accuracy and applicability to real-world security situations and digital analytics. https://jih.uobaghdad.edu.iq/index.php/j/article/view/4033Data MiningText MiningLatent Dirichlet Allocation (LDA) algorithmDigital ForensicsMachine Learning Techniques |
| spellingShingle | Shahlaa Mashhadani Detection the topics of Facebook posts using text mining with Latent Dirichlet Allocation (LDA) algorithm Ibn Al-Haitham Journal for Pure and Applied Sciences Data Mining Text Mining Latent Dirichlet Allocation (LDA) algorithm Digital Forensics Machine Learning Techniques |
| title | Detection the topics of Facebook posts using text mining with Latent Dirichlet Allocation (LDA) algorithm |
| title_full | Detection the topics of Facebook posts using text mining with Latent Dirichlet Allocation (LDA) algorithm |
| title_fullStr | Detection the topics of Facebook posts using text mining with Latent Dirichlet Allocation (LDA) algorithm |
| title_full_unstemmed | Detection the topics of Facebook posts using text mining with Latent Dirichlet Allocation (LDA) algorithm |
| title_short | Detection the topics of Facebook posts using text mining with Latent Dirichlet Allocation (LDA) algorithm |
| title_sort | detection the topics of facebook posts using text mining with latent dirichlet allocation lda algorithm |
| topic | Data Mining Text Mining Latent Dirichlet Allocation (LDA) algorithm Digital Forensics Machine Learning Techniques |
| url | https://jih.uobaghdad.edu.iq/index.php/j/article/view/4033 |
| work_keys_str_mv | AT shahlaamashhadani detectionthetopicsoffacebookpostsusingtextminingwithlatentdirichletallocationldaalgorithm |